EFX: Earnings Accelerated. Cash Has Not Caught Up.

The 60-second read

Five things to know

1. Operating demand remains broad. Diversified Markets grew 8% reported; all three reporting segments grew revenue despite mortgage weakness.
2. The headline EPS growth needs a bridge. GAAP EPS rose 38.8%; adjusted net income rose 5.7%, with a 4.8% smaller diluted share count supporting per-share growth.
3. The outlook softened. Full-year revenue midpoint fell $84M to $6.661B; adjusted EPS midpoint fell only four cents to $8.50 as management targets cost offsets.
4. Cash distributions exceed year-to-date free cash flow. Repurchases and dividends totaled $851.4M versus $763.2M of OCF less gross cash capital expenditures.
5. Our model leaves little immediate cushion. The $175 base value is close to the $175.17 intraday quote; the $130–230 scenario band reflects mortgage, cash conversion and valuation risk.

Three scenarios, one probability-weighted number

Analyst scenarios for the next twelve months; probabilities are judgments, not measured frequencies.

Signposts: what would change our mind

Thresholds are our monitoring rules, not management promises. This is our first EFX packet, so there are no prior Charged Alpha signposts to grade. Revisit the same definitions next quarter.

The tape and the price paid

At $175.17, the saved FMP quote showed a 10.47% advance from the previous close of $158.57. It was observed at 14:49:05 UTC on October 20, during trading. This is an early reaction, not a completed session return. The price can move materially before the close, so the model uses a timestamped snapshot rather than claiming that the market has delivered a final verdict. A strong bounce can coexist with a lower fundamental outlook when expectations were already depressed.

The available 52-week range was $137.01–238.94. The price is well below its high, but that alone does not establish cheapness: the relevant comparison is the cash shareholders can reasonably receive. Our model uses 117.1M shares outstanding from the September balance sheet, which differs from the provider's current share estimate. We show model equity value and provider market capitalization separately rather than silently mixing denominators. The stock is a US-listed common share, not an ADR, fund or bank.

Historical returns below are deliberately labeled close-to-close event-day windows. They can contain market moves and, for after-hours releases, may exclude the immediate post-release session. They therefore illustrate price volatility, not causal earnings surprise effects. Dates outside the saved fifteen-month history have no calculated return. Today's intraday observation is excluded from the historical completed-close series.

What the quarter delivered

Revenue of $1,692.5M exceeded the saved FMP consensus of $1,688.04M by only about $4.46M. Adjusted EPS of $2.26 exceeded $2.20 consensus by six cents. Those are positive surprises, but the revenue beat is small relative to the size of the business. The issuer's prior third-quarter range was $1.680–1.710B of revenue and $2.15–2.25 of adjusted EPS. Revenue landed within that band; EPS finished one cent above its high end. A beat versus analysts and a cut to the annual outlook are therefore both true.

Year-over-year revenue growth computed from reported dollars is 9.6%, rounded to 10% by management. Local-currency growth was 9%. The distinction matters because currency translation contributes to reported comparisons, especially outside the United States. Sequential revenue declined from $1,700.1M, a reminder that annual growth does not remove seasonality or mortgage sensitivity. We use comparable quarter growth to judge demand and reserve sequential comparisons for specific operating questions.

The new fourth-quarter range is $1.605–1.635B of revenue and $2.07–2.17 adjusted EPS. Full-year revenue guidance is $6.646–6.676B and adjusted EPS guidance is $8.45–8.55. The revenue midpoint is $84M below the prior $6.745B midpoint, while adjusted EPS midpoint is four cents lower. Management says weaker mortgage activity explains the revenue reduction and cost actions should recover much of its gross-profit impact. That is an execution commitment still to be tested, not a completed cash benefit.

Where the growth came from

Workforce Solutions generated $693.9M of revenue, up about 7%, and $306.3M of operating income. Verification Services contributed $603.2M, while Employer Services declined to $90.7M. That split is important: a broad segment label can conceal different end markets. Diversified Markets within Workforce grew 9%, with Talent Solutions above 20%, consumer lending up high double digits and government up mid-single digits. Workforce mortgage revenue was approximately flat. These are management's growth descriptions, not independently measured transaction counts.

USIS revenue rose to $599.1M from $530.2M, with $145.4M of operating income. Mortgage revenue grew 28% while diversified revenue grew 6%. Mortgage revenue can outperform the underlying lending market through product mix, pricing and content per transaction, so revenue growth should not be read as an equal increase in mortgage originations. The release does not provide enough detail to independently allocate that 28% among volume, price and mix. We do not manufacture that bridge.

International revenue was $399.5M, up 9% reported and 7% in local currency. Asia Pacific grew 20% reported versus 12% locally; Latin America grew 12% versus 8%. Those differences show why the currency-qualified result is the cleaner operating comparison. International adjusted EBITDA margin slipped to 30.9% from 31.3% despite growth. The pending Círculo de Crédito acquisition may expand the franchise, but its expected fourth-quarter close adds integration and capital-allocation questions. It is an ordinary business acquisition and does not change the identity of the listed parent covered here.

The earnings bridge

GAAP operating income improved by $91M. Cost of services increased by $89.2M and depreciation and amortization by $11.1M, while selling, general and administrative expenses declined by $43.7M. Prior-year resource-realignment costs were $43.9M and were absent from the current adjusted reconciliation. Thus, a substantial part of the headline GAAP improvement reflects a cleaner comparison, not a similarly sized increase in recurring operating efficiency. Adjusted EBITDA grew 11%, with margin expanding about 42 basis points, much less than the roughly 388-basis-point GAAP operating-margin increase.

The current income statement moves from $355.3M operating profit through $62.3M interest expense and $3.8M other income to $296.8M pretax profit. Taxes were $83.4M; consolidated net income was $213.4M, of which $1.7M belongs to noncontrolling interests. Parent net income therefore equals $211.7M. This is a useful check against data-provider fields: the primary statement's D&A is $194.4M, while a provider's standardized depreciation field differs. We use the primary line for this bridge.

Adjusted parent net income is $267.4M after adding acquisition amortization and other defined adjustments and subtracting their tax effects. It grew 5.7%. Dividing that income by the old 124.1M diluted shares would produce about $2.15 per share; dividing by 118.1M produces about $2.26. The smaller denominator contributes approximately eleven cents to the current per-share result. That is a real benefit to each remaining share, but it consumes capital and cannot be treated as free operating growth. Rounded reported EPS also means the exact ratio bridge differs slightly from the rounded percentage headlines.

Nine checks on earnings quality

Cash conversion is the most constructive counterargument to the skeptical earnings bridge. Nine-month free cash flow of $763.2M is about 100.3% of adjusted parent net income of $761.1M. Earnings are not simply disconnected from cash. The problem is the direction: OCF was nearly flat and gross capital expenditures rose, leaving free cash flow below the prior-year period. Both the conversion level and its growth matter. A high conversion ratio does not erase a weakening cash trend, and a modest cash decline does not by itself prove aggressive accounting.

Stock compensation was $74.5M in the nine-month cash-flow statement, about 1.48% of revenue. It is a noncash OCF add-back with a real ownership cost. Our valuation assumes the current share base rather than projecting endless shrinkage; this avoids crediting buybacks without funding them. The GAAP-to-adjusted gap is principally acquisition amortization, which may not require current cash but reflects capital previously paid for businesses. Investors can reasonably capitalize a sustainable cash franchise while still demanding that acquisitions earn an adequate return.

Receivables reached $1,104.2M. Dividing ending receivables by quarterly revenue and multiplying by 91 gives a 59.4-day proxy. This is not average-receivables DSO and is affected by mix and timing. Inventory is not a useful operating-quality measure for this data business. Tax expense is approximately 28.1% of pretax income, and net interest rose. We flag those rather than assuming a lower future tax rate or perpetual refinancing savings. No Beneish or Altman score is shown because blindly fitting standardized industrial formulas to partially normalized data would imply precision the evidence does not support.

The cash and balance-sheet test

At September 30, Equifax held $192.9M of cash against $5,206.7M of short- and long-term debt, giving $5,013.8M of net debt. That is approximately 2.39 times trailing adjusted EBITDA. It is not a net-cash business, and the equity does not have a large liquid-asset floor. The July financing extended the debt maturity mix: short-term debt fell substantially from June while long-term debt increased. Moving debt between buckets improves maturity management but is not itself deleveraging. We use total debt minus cash to avoid that optical trap.

Goodwill and identifiable intangible assets together represent about 67.6% of total assets. Their earning power matters more than liquidation book value, and impairments could affect reported equity without an immediate cash outflow. Conversely, recurring capitalized software investment requires cash even though it is not expensed immediately. That is why our cash model subtracts gross cash capex. The annual report also presents a net capital-expenditure KPI that differs from the cash-flow-statement line; the packet consistently uses the latter. Changing definitions between years would distort the apparent improvement.

Nine-month repurchases of $652M and dividends of $199.4M total $851.4M, exceeding free cash flow by $88.2M before excise taxes and other financing flows. This is not a liquidity crisis: quarterly cash generation is seasonal and the business has access to financing. But distributions above cash generation cannot continue indefinitely without raising leverage or using cash balances. The latest quarter's $92M buyback was smaller than earlier spending. With an acquisition expected, the next balance sheet should demonstrate a deliberate allocation among growth investment, shareholder returns and financial flexibility.

Three routes and the price-implied case

Our base fair value is $175 per share. The first route applies 22 times the company's $8.50 adjusted EPS midpoint, yielding $187. The multiple is an analyst judgment that recognizes recurring data products and growth while discounting leverage, mortgage exposure and non-GAAP adjustments. It is not a peer median presented as fact. The second route uses 12 times $2,097.9M of trailing adjusted EBITDA, then subtracts $5,013.8M net debt, $120.8M redeemable noncontrolling interests and $20.5M ordinary noncontrolling equity. Dividing by 117.1M common shares produces $170.96. Including both minority claims makes this route conservative and explicit.

The third route begins with $1,104M of trailing cash flow after gross capex, derived from FY2025 plus current nine months minus prior nine months. We grow it 8% annually for five years, discount at 10%, and use 3% terminal growth. The present value of the five explicit cash flows is $5,226.1M; terminal value contributes $14,820.5M in present terms. Their sum divided by 117.1M shares gives $171.19. This is an equity-cash proxy after interest, assuming neutral incremental borrowing. Debt is not subtracted a second time. It is not an unlevered enterprise DCF, and it does not assume additional buybacks magically reduce shares.

The three routes average $176.38, rounded to $175. They share economic inputs, so agreement is not three independent proofs. Terminal cash contributes heavily to the discounted model; small changes in long-run growth or required return can move value substantially. At the observed quote, the model enterprise value is approximately $25,667.5M, or 12.23 times trailing adjusted EBITDA. The cash model requires roughly 8.56% five-year growth to justify the quote with the other assumptions unchanged. That demand is plausible but leaves limited protection if cash growth lags.

High uncertainty warrants a 15% margin of safety in our framework, equivalent to $148.75 against the base value, subject to the thesis remaining intact. This is a valuation threshold, not a price forecast or instruction to transact. The $130 bear and $230 bull cases capture earnings and multiple risk. Their weighted $177.50 value is close to the base, so probability arithmetic does not rescue a strong upside case at today's snapshot. A better future outlook can raise value; a cheaper quote alone cannot repair a deteriorating business.

Wall Street context

The saved provider consensus was Buy, with 23 buy, 12 hold and one sell recommendation. Its average target was $202.31, median $208 and range $148–245. The average implies about 15.5% upside from the snapshot, compared with essentially flat value in our base case. These aggregates combine research published at different times; they should not be described as a unanimous response to this morning's release. The latest named records available in the pull predate the print.

Mizuho reduced its target to $185 from $210 on October 19, RBC Capital reduced its target to $177 from $194 on October 16, and Goldman Sachs reduced its target to $148 from $171 on October 2, according to FMP's recorded price-target news. We use those dates to expose freshness rather than treating the highest target as the most informed one. The underlying analyst reports were not supplied, so we do not attribute a detailed investment thesis to any firm beyond the recorded target change.

The same snapshot carried FY2026 revenue consensus near $6.721B, above the new $6.661B company midpoint, and EPS around $8.50. The following year's EPS estimate was approximately $9.84 from 15 analysts. Consensus can update after a release and may use adjusted definitions. We show it as a snapshot and compare like-for-like cautiously; it is neither a management commitment nor guaranteed growth. Our bull case of $10 EPS is close to that direction but deliberately remains an assumption with a valuation multiple attached.

Management commitments to grade

The strongest management evidence is not a slogan about artificial intelligence. It is broad diversified revenue growth, a new-product Vitality Index above 17%, and the ability to preserve much of the EPS outlook despite lower expected mortgage revenue. Those results suggest the cloud and data investments are producing commercial benefits. However, the Vitality Index is a product-contribution KPI; it does not independently identify AI revenue, incremental gross profit or realized cost savings. We do not translate it into an unreported AI profit figure.

Management's $150M AI productivity objective should be judged by expense and cash evidence over its stated strategic period. A goal cannot be added wholesale to earnings while the valuation already assumes higher cash growth. That would count the same operational improvement twice. The Q3 adjusted EBITDA margin improvement is encouraging but modest, and International's margin moved in the opposite direction. We would look for repeated consolidated improvement with no deterioration in customer service, data quality or compliance spending.

The capital-allocation score is mixed. Repurchases reduce the share count and amplify EPS, while the debt burden and acquisition pipeline constrain flexibility. The annual report's $3B authorization began in April2025; its December2025 remaining amount is a historical balance, not an assertion of current purchasing capacity. Future buyback execution should be assessed against intrinsic value and cash availability. A repurchase can improve EPS while destroying value if management pays too much, so share shrinkage alone is not a sufficient scorecard.

1. How much of mortgage revenue growth comes from price, mix and volume?
2. Which AI savings have been realized in cash rather than targeted?
3. What will the Círculo acquisition contribute after integration costs and financing?
4. What cash-generation threshold governs repurchases while leverage remains material?
5. Can the company reconcile Vitality Index growth with net customer economics and product profitability?

Ranked risks and counterarguments

The bear case is a prolonged mortgage slowdown that makes the reduced guidance another waypoint rather than a floor. Revenue is not entirely volume-driven, but a persistently weaker market can pressure growth, pricing tolerance and operating leverage. If cost offsets are delayed, the current premium to slower-growing businesses becomes harder to defend. Our model would fall through both lower earnings and a lower multiple, so the sensitivity is nonlinear in a difficult scenario.

The strongest bull response is that Equifax has already grown through a weak mortgage backdrop. Proprietary employment and credit data can be difficult to replicate, and new products can raise revenue per interaction. Workforce diversified growth and USIS mortgage outperformance support that argument. Historical free cash flow improved materially from2023 through2025 as cash capital investment fell. A return to that cash trajectory, combined with disciplined distributions, could justify a higher base than we use today.

Cybersecurity, privacy rules, data access and accuracy remain structural risks for a business built around sensitive information. The2017 incident still has small current reconciliation items; a small current accrual does not cap the potential cost of future events. Legal and regulatory costs may recur even when adjusted metrics exclude individual charges. International currency and integration risk add uncertainty, while debt means equity bears amplified changes in enterprise value. None of those risks is a prediction of failure; they explain why our valuation requires cash evidence and a margin of safety.

What changes the call

The next catalyst is the fourth-quarter outcome against management's newly narrowed ranges. We would not upgrade merely because adjusted EPS beats a lowered consensus by a few cents. The more useful combination is revenue at least inside guidance, sustained diversified growth, improving cash generation and leverage that remains controlled after any acquisition. Those observations would help distinguish durable efficiency from a temporary comparison benefit.

The expected fourth-quarter Círculo closing creates another checkpoint. The release identifies an expected timing, not an accomplished transaction, and we do not add unverified acquisition earnings to the base. At closing, purchase price, financing, acquired revenue, integration spending and any revised guidance should be reconciled. A deal can strengthen data coverage and still reduce near-term equity cash if the price or financing is demanding.

A change in mortgage conditions can act before the next quarterly report, but higher revenue from that segment must be separated from volume recovery. Longer term, a track record of realized productivity gains matters more than expanding AI terminology. The scheduled review dates below are analytical checkpoints; earnings dates beyond the current release were not confirmed in the saved primary evidence. Our thresholds should survive ordinary volatility and be revisited only when new facts change the economic case.

Financial appendix and methods

All dollar amounts in financial tables are US dollars in millions unless labeled per share. Quarter-end balance-sheet shares and weighted-average diluted shares serve different purposes: the first allocates enterprise value today, while the second reconciles reported quarterly EPS. We retain both instead of forcing them to match. Growth, margins, cash flow, leverage, valuation routes, scenario weights and the reverse cash-growth solve are computed in the saved derivation script. Rounded display figures can differ slightly from arithmetic performed on reported EPS.

Primary sources are Equifax's current8-K Exhibit99.1, the preceding eight earnings releases and the2025 and2024 annual reports. Current release tables order2026 before2025; the balance sheet compares September30,2026 with December31,2025; cash flows cover nine months. Reading those headers prevents the common error of treating year-to-date cash as a single quarter or comparing a quarter to an annual balance. SEC source files and hashes are retained with this packet's dataset. FMP provides market prices, standardized historical financials, consensus and dated target records. Where its standardized current fields differ from primary statements, the primary statement governs our display.

The historical financial summary below is intentionally compact; the retained dataset contains the complete provider income, balance-sheet and cash-flow fields and primary release tables for each quarter, including non-GAAP adjustments and segment detail. Historical employee counts not verified in the source review are not invented or used. No share-class voting premium, VIE discount, customer-concentration percentage or acquisition synergy is assumed without supporting evidence. We did not use a post-release conference-call transcript, and same-day consensus may lag the release. These limits affect precision, not the verified existence of the earnings print.

Free cash flow here means OCF less gross cash capital expenditure from the cash-flow statement. It is a non-GAAP analytical measure, excludes acquisitions and distributions, and is not automatically the amount available for buybacks after every obligation. EBITDA is also non-GAAP and omits cash interest and investment needs. The valuation uses both as cross-checks rather than treating either as distributable profit. Scenario probabilities and required return are explicit author judgments. This research is designed to make those judgments inspectable and falsifiable in the next update.